INTRODUCTION

It was 2012, and the story ricocheted around the world: Seneca faculty and students had built an operating system for a $35 computer that would put affordable technology in the hands of millions. And to paraphrase the old joke: after 10 years of hard work, Seneca’s Centre for Development of Open Technology, or CDOT, was an overnight success.

Figure 1
Chris Tyler and the prototype Raspberry Pi; a closeup of the Raspberry Pi 1, Model B

It was an amazing story. Led by Professor Chris Tyler, students and faculty from Seneca had developed a basic operating system based on Linux Fedora for the ultra-inexpensive Raspberry Pi microcomputer. But it was far from CDOT’s first contribution to the open source community. By 2012, the Centre’s partnership with the Mozilla Foundation, developer of the Firefox browser, was firmly established and CDOT had become an invaluable source of innovations and improvements benefitting millions of Internet users around the world.

To this day, CDOT continues to be a model of polytechnic-industry collaboration. Forging close ties between the worlds of academics, community, and industry, the Centre brings together students, faculty, developers, and experts to leverage open source technology for both public and commercial use.

Along the way, CDOT has become an important example of the capacity for the polytechnic system to foster innovation in teaching and learning, applied research and global community outreach. It has had a sustained and meaningful impact on the student experience, graduate career success, and faculty development. Its products and projects have benefitted a broad range of industry and non-profit partners. Hundreds of millions of people around the world touch contributions from CDOT every day when they use their web browsers and mobile devices.
The evolution of CDOT had three distinct phases: its informal beginnings in the 1990s; the 2002–09 period that saw increasing prominence in the open source community; and the last decade, when significant funding and national recognition for its leaders gave the Centre’s work an important, stable foundation.

What has not changed is CDOT’s relentless commitment to students and to understanding the power of connecting real-world experience with sophisticated theoretical knowledge.

THE EARLY YEARS: 1990-2002

In the early 1990s, computer curricula in post-secondary education reflected industry practice of the time. Dumb terminals were used to access central minicomputers and mainframes running proprietary software systems. Personal computers—generally IBM or Microsoft DOS/Windows–based—were used mostly to run stand-alone office applications and custom software, with limited connectivity through proprietary networks. The Internet had only just become a publicly accessible resource.

The origin of the open source movement is anecdotally attributed to events surrounding a mundane faculty problem with printer jams at MIT in the early 1980s. Faculty had solved a pesky issue with jams by inserting their own “hacker code” in the network to notify each other when a jam occurred. They could then address it before their print queue was affected. When a new laser printer was installed in the computer studies department, faculty were not given access to its proprietary code. They reacted by declaring their intent to clone a version of UNIX (a proprietary operating system developed by AT&T and widely licensed to other vendors), eventually calling it GNU (a pun of an acronym, standing for Gnu’s Not Unix), to give complete freedom to users, with the stipulation that they could use, copy, distribute, and modify the software as long as they shared the modified source code. This initiative is reputed to be the foundation for, writes Neary (2018), “the decentralized, collaborative development model that the free software movement became” (para 9.).

By the mid-1990s, the Linux operating system, a full-featured UNIX clone that borrowed many elements from the GNU project, was stable and being adopted by major players in enterprise computing (Red Hat, 1999). It was developed in a way that was completely different from other operating systems. It was open source software with the source code freely available to all users. Linux benefitted from a global community of highly experienced programmers and academics constantly enhancing it.

While there are various definitions of open source software, the common element is source code that is available to anyone to explore or change. As the website opensource.com (n.d.) notes:

The term originated in the context of software development to designate a specific approach to creating computer programs. Today, however, “open source” designates a broader set of values—what we call “the open source way.” Open source projects, products, or initiatives embrace and celebrate principles of open exchange, collaborative participation, rapid prototyping, transparency, meritocracy, and community-oriented development. (para 2)

The open source movement, with its ethical hacking ethos, even had its own book: The Cathedral and the Bazaar, by Eric S. Raymond. This essay served as justification to support the release of the source code for the Netscape browser and the establishment of the Mozilla project in 1998. This was the first major release of an existing widely available desktop software as free software (Kornblum, 1988). It set the stage for the expansion of open source curriculum at Seneca and for the establishment of CDOT itself. The Mozilla project’s leadership positioning in this space was cemented by the release of the Firefox browser in 2004, making it an obvious choice of partner for CDOT—even though it would have been hard at the time to envision the impact this alliance would have on both organizations in the future.

But this relationship, as obvious as it seems now, was not yet even a twinkle in either partner’s eye. That was several years away.

In the meantime, Seneca’s School of Computer Studies had moved into a former IBM training centre, and by the mid-1990s, open source curriculum was introduced by Professor John Selmys and others such as Evan Weaver (later chair of the school) as the department started to use Linux to teach UNIX-related material with cheaper PC clone technology in place of more expensive minicomputers.

With the growth of the school, in 1999, the programming programs moved to the new Seneca@York Campus on the grounds of York University. A group of faculty coalesced into a dynamic informal working group of open source enthusiasts and collaborated extensively. Student engagement was supported by “Linux Installfests” to help students install the open source operating system on their personal machines. This gave students the ability to run and administer Linux servers at home, at no cost, from the first semester of their program—a head start that accelerated their hands-on experience.

Figure 2
Linux Cluster at Don Mills Campus c. 1995

By 2002 three key components were in place:

1. Faculty leaders with a commitment to fully engaging with the open source community and supporting the nascent industry adoption of open source software. They aspired to become internationally recognized contributors to open source software development and to involve their students in this experience.
2. A faculty team creating and delivering unmatched depth and breadth in higher education open source curriculum in programming and networking programs.
3. Academic leadership that provided necessary equipment and the necessary freedom to use it as required to support open source applications and their development.

CDOT TAKES FORM: 2002-09

The Centre for Development of Open Technology began in 2002 with the group of Seneca faculty from the School of Computer Studies (now the School of Information and Communications Technology) deciding to formally support and promote open source technology. The thriving and energetic collaboration involved a large number of Seneca faculty teaching open source courses while many also worked on passion projects that ran the gamut from releasing software written as open source to helping other institutions, such as local high schools, leverage the benefits of open source technology.

The choice of “open technology” rather than “open source” was both deliberate and forward-looking. It proved prescient in providing scope for the future expansion of the open source approach from software code to technologies across disciplines.

The Centre was also located in a highly visible space, with bold branding in a main traffic area of the TEL (Technology Enhanced Learning) building, a facility shared with York University. This flagship location grounded the CDOT initiative with a degree of prominence and permanence. Mozilla’s Mike Shaver writes: “The space allocated by administrators served as an expression of values” (personal communication, June 13, 2018).

As is true for all open source initiatives, a grassroots, bottom-up approach defined the initiatives at Seneca. Professor David Humphrey joined Seneca in 1999 as an expert user of open source software. But he had bigger plans: he aspired to be a recognized and respected contributor, and mapped out a strategic pathway to achieve that goal. Although the open source community allows for anyone to contribute code, the ability to participate at scale with meaningful contributions requires extensive collaboration. Humphrey writes: “The stereotype of a lone geek writing lines of code in his parents’ basement is only relevant to very minor projects involving 200 lines of code … to contribute to major projects with thousands or millions of lines of code, collaboration with top tier programmers is a necessity—and to interact with these individuals a programmer must first establish their credibility” (personal communication, June 1, 2018).

Humphrey’s aspiration to reach that goal proved to be fundamental to the success of CDOT. His vision helped the team conceive of a “circle of virtue” where code, curriculum, and community involvement resulted in the credibility needed to attract opportunities to make significant contributions to high-profile projects that would inspire and inform the activities of students, faculty, graduates, and industry partners alike.

Key to that virtuous circle was developing credentials to match the ambition. Open source curriculum development continued with an expanded set of diploma course offerings in programming, networking, and system administration. The growing popularity of Linux created a need for Linux expertise in the corporate environment. The Internet Systems Administration, a Linux-based graduate certificate, was added in 2000. The program specifically intended to produce Linux system administrators.

An important moment came with the provincial government’s change in policy that allowed colleges to offer four-year bachelor’s degrees and at the same time designated Seneca as a differentiated institution that could offer a higher number of degrees. By 2002, Seneca received consent to offer the Bachelor of Software Development degree, and the first intake of students came a year later.

Along with granting degrees, polytechnics such as Seneca started building capacity to formally engage in applied research. A founding member of Polytechnics Canada and the Colleges Ontario Network for Industry Innovation, Seneca collaborated with like-minded institutions to combine expertise and capacity as well as advocate with both federal and provincial governments for applied research funding.

But the partnership that proved seminal in CDOT’s early years was with the Mozilla Foundation. The foundation had evolved from the informal structure spawned by the release of the source code for the Netscape browser suite in the late 1990s. That release set off enormous energy in the open source community, quickly expanding the focus from improving the browser to a range of other projects. The common bond was a commitment to creating free software.

In 2004 the Mozilla Foundation released the first version of the Firefox browser, and in less than a year it was downloaded over 100 million times. According to Mozilla’s website (n.d.): “New versions of Firefox have come out regularly since then and keep setting new records. The popularity of Firefox has helped bring choice back to users. The renewed competition has accelerated innovation and improved the Internet for everyone” (para. 5).

Nesbitt (2014) writes:

Around 2005, David Humphrey, a faculty member who was part of the self-declared CDOT team, contacted Mozilla for some help with the Firefox code base, as he wanted to experiment with the user interface as part of an applied research project he was supervising to help a local inventor with product development. Mike Shaver, Chief Evangelist and VP of Engineering from Mozilla, was interested enough with Humphrey’s experiments that he offered Mozilla’s help at no charge. Once Shaver saw what David’s student researchers were capable of doing, he brought some Mozilla work to Humphrey, and the rest, as they say, is history. (para. 5)

Shaver vividly recalls his first interaction with CDOT:

David Humphrey’s class in the (software development) degree program was doing work with a local company on touch screen interfaces. When David reached out to Mozilla, I happened to be in Toronto, and checked it out. I was impressed with the energy, thoughtfulness, and value of the open source opportunity for students to have industry experience. I saw an incredible commitment to this student experience—connecting students to parts of source code, to software engineers, and how to interact with them as professionals. (personal communication, June 6, 2018)

The relationship with Mozilla opened the doors for CDOT to work on large-scale projects, using infrastructure in place because of Seneca’s long history in high-performance computing.

CDOT undertook a wide range of projects that serve as early exemplars of applied research activities in the college system and supported the necessary capacity-building for applied research projects at Seneca at scale. Along with the relationship with Mozilla, companies such as IBM and Red Hat were providing funding and, as important, opportunities for senior students to work on their capstone degree projects. As a result, students and faculty have been involved in large-scale projects, developing the capacity to build prototypes, software solutions, or product and process innovations.

David Humphrey recounted the contribution of Andrew Smith, one of his first students in Seneca’s software design degree program. As a student, Smith spent a year writing the code for a new web standard for animated images called APNG, and it shipped with Mozilla’s Firefox browser in 2007. “Two years ago Apple implemented it in Safari and iOS. And in 2018, 10 years after Andrew, Google finished their work to add it to Chrome and Android. Between Google, Apple and Mozilla, the APNG standard affects just about every device that runs the web. That’s literally billions of devices worldwide” (D. Humphrey, personal communication, July 24, 2018). Andrew Smith now teaches at CDOT.

When it came time to apply for a large, multi-year applied research grant in 2009 through the Natural Science and Engineering Research Council (NSERC), the track record was formidable, writes Evan Weaver in his NSERC CCIP program application:

Seneca (CDOT) became an increasingly active participant in open source research and development with some of the largest open source projects in the world. This engaged students from the School of Computer Studies in real development activities, resulting in their work being used by hundreds of millions of people on a daily basis. By 2009 over 200 students had experienced hands-on open source development projects with Mozilla (Firefox), IBM (Eclipse) and Red Hat (Fedora) through Seneca’s open source courses. (Applied research in open source software: Leveraging community collaboration for competitive advantage, November 23, 2009)

The list of projects under the CDOT banner spanned the open source world—everything from software for a next-generation human interface device to an open source search engine to compiler code.

Faculty realized credibility with the developer community was essential to contribute to it at a meaningful level. Open source evangelism in the user community and networking at a senior level with leaders in the developer community were necessary precursors to achieving this goal.

One of the early targets was high schools. Professor John Selmys led the first project at Emery Collegiate in Toronto in 2003. Seneca donated servers, networking equipment, and client machines to build a “thin client lab” running the Linux Terminal Server Project. A thin client is a networked computer that runs without a hard storage disk on the user’s machine offering a lower cost solution than thick clients requiring hard disks at every desktop. Seneca students from the Linux-based Internet Systems Administration graduate certificate program configured and set up the lab and trained the teachers and administration on how to use the software.

Other projects followed, including a 2006 Wi-Fi camp for high schools led by professors Danny Roy and John Selmys with teams of students and teachers from across the province. Students built their own “antennae” using open technologies and used them in a competition to find hidden access points.

The group shared their enthusiasm for open source systems widely within the Seneca community. They distributed test-drive Linux discs to all Seneca staff—with an auto-run program to install Firefox and Open Office on Windows.

The Linux club provided a way for students interested in open source to meet and a lab to experiment with hardware and software. The club’s main activity was running Linux Installfests two or three times a year to help any Seneca student or faculty get working versions of the Linux OS on their personal computers.

Another critical step in the establishment of CDOT was what would become known as FSOSS, the Free Software and Open Source Symposium. Starting with a small gathering in 2002, the symposium quickly expanded and became recognized as one of Canada’s best events for bringing together developers and educators working with open source. Early participants and sponsors included software developers Novell and Sun Microsystems.

By 2007, the conference had three tracks, with industry speakers from Microsoft, Facebook, Mozilla, IBM, and SourceForge.net. Best practices for teaching with open source were presented by participants from as far as Transylvania and South Africa, and the keynote was delivered by a member of the Open Source Research Group at SAP Labs, Palo Alto, California. Sponsors included the Mozilla Corporation, BMO, Novell, Amsdell, and Apple.

Anchoring the development of CDOT was the active and total engagement of students. As faculty members and active members of CDOT, David Humphrey and Chris Tyler saw the potential of bringing real problems into the classroom and having students work on open source solutions (Nesbitt, 2014).

The interplay between the meaningful contributions being offered to industry partners in conjunction with the developer community and software company sponsors, and its incorporation into curriculum, giving students the ability to contribute to real projects as part of the real process, led to the crystallization of CDOT’s mission and the development of a new model of instruction and collaboration. Faculty and students specialized in the technologies and methodologies of open source, which became one of the dominant models of software development by the end of the decade. The approach was articulated in an unpublished paper by David Humphrey:

The CDOT model was built on three foundations:

1. Do It: We get our faculty directly embedded within large open source projects, working on open technology and standards; this is our main avenue for influence within the industry, as well as how many of our graduates have been able to create their own jobs at influential companies.

2. Teach It: We actively engage our students with open source software throughout their programs, and specifically use capstone courses and professional options to get them involved in real-world open source projects, engaging directly with the community.

3. Apply It: We work with industry partners to leverage our knowledge of key open source technologies in order to create research and development opportunities that lead to innovation and commercialization.

Figure 3
The CDOT Model

David Humphrey’s work on the Audio Data API was a particularly important project for its role in the invention of a way to programmatically generate sound on the web. Shipping with Firefox version 4 in September 2010, it paved the way for the standardization of Web Audio that was later shipped in most web browsers. It was a particularly notable example of the teaching philosophy adopted by CDOT faculty. Humphrey (2013) writes:

We often talk about the web as a democratizing technology and as an open platform for everyone. However, it’s not just that wikis are open, or that you can make your own web page. The web platform itself is open and malleable, and Mozilla is willing to enable a community of participation all the way down the stack. We didn’t make a web page; we remade the browser, and made a whole new kind of web page possible. There aren’t many other platforms where that’s even possible. The web is a different sort of platform, and Mozilla a different sort of gatekeeper … ” (para. 6)

As Mike Shaver and Mozilla became more involved with CDOT, they were quick to identify unexpected benefits of participating in this model. Not only did the Mozilla project gain from the contribution of coding efforts on the part of Seneca’s students, they reaped the benefit of the “Do It. Teach It. Apply It.” approach and embedded it within their own business operations. Software engineers from Mozilla participated as guest lecturers and mentors. They identified and developed leaders within their organization. Shaver writes: “[T]hese mentors learned how to make choices, how to step in, when to step in by working with Seneca students. Learning from Seneca has become structural at Mozilla, how we onboard people, how we communicate.… Mozilla got a lot of out it” (personal communication, June 6, 2018).

The Mozilla foundation was understandably keen to replicate the CDOT model at the top-tier post-secondary institutions they partnered with in Canada and the US. Shaver writes: “Once CDOT became a thing and it was clear Seneca was a real asset to Mozilla and this approach was working, top tier colleges in US wanted to do the same thing but couldn’t find traction. The attempts to replicate the model failed. The critical component was missing at other institutions—the dedication and prioritization of the student experience was what drove CDOT. Vertical support for this was provided by administrators. David Humphrey himself was infectious, secret weapon.” As a key success factor, “the energy of the lead faculty dwarfed the institutional resources” (personal communication, June 6, 2018).

NATIONAL FUNDING AND RECOGNITION: 2010 – PRESENT DAY

The capacity for applied research at Canadian colleges, and Seneca’s CDOT in particular, was formally recognized and supported in the second round of NSERC grants awarded in 2010. The College and Community Innovation Program focused on projects benefitting local companies, particularly small and medium-sized enterprises, and Seneca received five-year, $2.5 million funding.

Consistent with the CDOT approach, Weaver’s 2009 NSERC CCIP program application started with students: “To provide more opportunities for our students to be involved in innovative, large scale, open source software development activities. Students will have increased access to academic and paid research positions that result in contributions that are relevant and valuable to external partners. This will increase the number of graduates who have the skills and experience needed to work with the large code bases typical of modern software development.”

The project’s research area spanned a range of open source opportunities that would benefit small and medium-sized enterprises, including extending the open web, incorporating open source into product development, improving enterprise deployment of open source, and improving open source tools and processes.

This support gave CDOT its first stable external funding. Mike Shaver describes the importance of this grant “as an existential moment for CDOT. The application process was new to Mozilla and much more complex than expected. We learned what it meant to be grant partners; what was the commitment, what were the potential risks—the financial implications. It took a lot of explaining across the organization. However, the value of this funding and the credibility it would bring to the college made the effort worthwhile” (personal communication, June 6, 2018).

That credibility was further enhanced when Chris Tyler, whose pioneering work with Red Hat and the Fedora operating system was an early CDOT success, was named the NSERC Industrial Research Chair for Colleges in Open Source Technology for Emerging Platforms in 2012. Backed by a $1 million renewable five-year grant, Tyler focused on foundational software for new energy-efficient computing platforms, ranging from the revolutionary $35 Raspberry Pi computer to large data centres, with industry partner Red Hat Canada. His focus was on ARM computer systems, which have the potential to reduce data centre energy, space, and cooling requirements by 90% or more. The grant was renewed in 2017. Building on the research into ARM computer systems, Tyler now focuses on related areas of super-embedded computing and development practices.

Two years after Tyler was awarded the industrial research chair, David Humphrey was the first college faculty member to receive NSERC’s Synergy Award for Innovation, recognizing the long and fruitful partnership between CDOT and Mozilla.

Figure 4
David Humphrey and former Governor General David Johnston

Surman (n.d.) writes of the CDOT-Mozilla collaboration:

The web developer’s open source platform has made it possible for hundreds of Seneca students to contribute their own ideas and code to help make Firefox more interactive in the online media and gaming space.…A dozen Seneca graduates now work for Mozilla and are mentoring the next generation of Seneca students. The success of the partnership has helped Mozilla grow its Canadian staff significantly: Since they began working together in 2005, Mozilla’s Toronto office has grown from two employees to more than 80. (para. 3-4)

According to Seneca’s (n.d.) website:

Ever since the Mozilla Foundation released Firefox 4, new releases of Firefox have featured groundbreaking technology developed by Seneca students, graduates, and faculty. As a result of Seneca’s on-going partnership with the Mozilla Foundation, CDOT played an integral role in the creation of key features in various releases of Firefox. The team developed the world’s first web audio API, implemented the much requested web-gaming feature known as Mouse/Pointer Lock for 3D interfaces, implemented captioning for web video in the form of WebVTT, and worked on countless bug fixes and other features. (para. 2)

Many of these features spawned at CDOT are not just found on the Firefox browser, but have also been adopted by Microsoft in Edge, Google in Chrome, and Apple in Safari.
Student success has always been the driving force behind CDOT. Shaver writes:

Through this hands-on applied approach to curriculum and research, we have produced graduates with a mix of deep skills and real-world experience.… This is a result of the deep engagement our faculty have with open source.

Through open source collaboration, students develop communication skills and resilience that are far ahead of students learning in a theme-park stay-on-the-rails education. The experience students have brings breadth. It’s analogous to the role Liberal Arts plays. Some students find it off-putting at first—these are projects, not homework and the software might not work.

Being able to participate effectively in the populist, broadening community that supports the Mozilla Foundation is critical. We will be here at Seneca long term to train and hire. It’s an important talent pipeline. Seneca’s open source programs serve as a year-long audition for our team. Even students involved in smaller projects were able to join us as interns, and summer students (were) way ahead of the curve understanding release engineering and DevOps. They already knew how things worked” (personal communication, June 6, 2018)

CDOT continues to evolve, but a constant remains the faculty-driven “Do It. Teach It. Apply It.” model, incorporating real-world student projects feeding into a global open source community. Deep industry partnerships mean a steady flow of new projects—from raw development to fixes and patches—to challenge each new intake of students.

It’s the polytechnic applied research model in action. And it continues to be an overnight success, well into its second decade.

POSTSCRIPT

On June 5, 2018 Microsoft announced it was purchasing GitHub — a website supporting open source developers to manage versioning of their files — for US$7.5 billion.

Observed David Humphrey:

GitHub is 10 years old, and reflects the growth and dominance of open source in the same period that CDOT has been most active. Open Source is now so mainstream that it makes sense for one of the world’s largest tech companies to spend $7.5 billion in order to align itself. Long gone are the days of loners in their basements working on code for free. Open Source is now how we all do software, how we all do business. Overall, it’s a current detail that helps to add support to the long-term bet we made on Open Source at Seneca with CDOT.

References